effort-estimator

name	effort-estimator
description	Analyze feature requests and provide technical effort estimates with risk assessment
level	2

Effort Estimator Skill

Note: This is a sample skill that estimates technical effort for feature requests. Your version should be tailored to your team's velocity and complexity factors.

Purpose

This skill analyzes feature descriptions (written in product/business language) and provides:

• Effort estimates with confidence levels
• Breakdown by discipline (backend, frontend, testing, etc.)
• Risk identification and complexity factors
• Recommendations for breaking down large features

When to Apply

Use this skill when:

• Product managers describe a new feature
• Stakeholders ask "How long will this take?"
• Planning sprint capacity
• Prioritizing backlog items by effort vs. value

Effort Categories

🟢 Small (< 3 days, 1-5 story points)

• Single component or endpoint change
• No database schema changes
• No new dependencies
• Existing patterns can be reused
• Limited testing scope

Example: Add social media share buttons to character pages

🟡 Medium (3-7 days, 8-13 story points)

• Multiple component changes
• Minor database changes (new columns, simple migrations)
• Some new patterns or integrations
• Moderate testing requirements
• May touch 2-3 layers (e.g., backend + frontend)

Example: Add episode filtering by genre on the browse page

🟠 Large (1-2 weeks, 21+ story points)

• New feature area
• Database schema changes with migrations
• New API endpoints
• New frontend pages or complex components
• Comprehensive testing required
• Multiple team members involved

Example: Episode Detail pages with transcripts, character appearances, and streaming links

🔴 Extra Large (> 2 weeks, requires breakdown)

• New subsystem or service
• Significant infrastructure changes
• External service integrations
• Performance or scaling implications
• Should be broken into smaller deliverables

Example: Full-text search across all content with autocomplete

Complexity Factors

When estimating effort, consider these complexity multipliers:

Data Complexity

Factor	Impact
New table(s) required	+50% effort
Many-to-many relationships	+30% effort
Data migration needed	+25% effort
External data sources	+40% effort

Integration Complexity

Factor	Impact
New external API	+50% effort
Authentication changes	+30% effort
Real-time updates needed	+40% effort
Third-party service	+25% effort

Frontend Complexity

Factor	Impact
Complex state management	+25% effort
Animations/transitions	+20% effort
Offline support	+40% effort
Accessibility requirements	+15% effort

Testing Complexity

Factor	Impact
E2E testing required	+20% effort
Performance testing	+25% effort
Security testing	+30% effort
Edge case coverage	+15% effort

Estimation Framework

When analyzing a feature request:

Step 1: Identify Components

Reference docs/ARCHITECTURE.md to understand:

• Which backend routes need changes
• Which frontend pages/components are affected
• Database table impacts
• API contract changes

Step 2: Break Down by Layer

Feature: [Feature Name]

Backend:
- Routes: [list changes needed]
- Services: [business logic changes]
- Database: [schema changes, migrations]
- Estimate: [Small/Medium/Large]

Frontend:
- Pages: [new or modified pages]
- Components: [new or modified components]
- State: [state management changes]
- Estimate: [Small/Medium/Large]

Testing:
- Unit tests: [scope]
- Integration tests: [scope]
- E2E tests: [if applicable]
- Estimate: [Small/Medium/Large]

Infrastructure:
- Changes needed: [if any]
- Estimate: [None/Small/Medium]

Step 3: Identify Risks

Risk	Likelihood	Impact	Mitigation
[Risk name]	High/Medium/Low	High/Medium/Low	[How to reduce]

Step 4: Output Estimate

## Effort Estimate: [Feature Name]

**Overall**: [Small/Medium/Large/Extra Large]
**Confidence**: [High/Medium/Low]
**Story Points**: [Range, e.g., 13-21]
**Calendar Time**: [Range, e.g., 1-2 weeks]

### Breakdown
| Area | Effort | Notes |
|------|--------|-------|
| Backend | [Small/Medium] | [Key work] |
| Frontend | [Small/Medium] | [Key work] |
| Testing | [Small/Medium] | [Scope] |
| Infrastructure | [None/Small] | [If needed] |

### Risks
- [Risk 1]: [Mitigation]
- [Risk 2]: [Mitigation]

### Recommendations
- [Break down suggestion if Large/XL]
- [Dependencies to resolve first]
- [Spikes or research needed]

Example Estimates

Example: Episode Detail Pages

Feature: "Users want to click on an episode and see full details including synopsis, character appearances, memorable quotes, and streaming links."

Analysis:

• Backend: New /episodes/:id/full endpoint with joined data (Medium)
• Frontend: New EpisodeDetail page, multiple components (Medium)
• Database: No schema changes, but complex joins (Low complexity)
• Testing: Component tests, API tests, E2E for navigation (Medium)

Overall: Large (1-2 weeks) Confidence: High (similar to Character Detail) Story Points: 21-34 Risks:

• Many-to-many relationships (characters in episodes) add complexity
• Streaming links may require external API integration

Example: Full-Text Search

Feature: "Users want to search for anything - characters, quotes, episodes - from a single search box."

Analysis:

• Backend: New search service, indexing strategy, query optimization (Large)
• Frontend: Search bar, results page, autocomplete, filters (Medium)
• Database: May need search index (PostgreSQL full-text or external) (Large)
• Infrastructure: Consider dedicated search service at scale (Medium)

Overall: Extra Large (> 2 weeks, break down required) Confidence: Medium (depends on search technology choice) Story Points: 55-89 Recommendations:

• Spike: Evaluate PostgreSQL full-text vs. Elasticsearch
• Phase 1: Basic search without autocomplete
• Phase 2: Autocomplete and advanced filtering
• Phase 3: Search relevance tuning

Confidence Levels

• High: Feature is similar to something we've built before
• Medium: Some unknowns, but architecture is understood
• Low: Significant unknowns, new technology, or unclear requirements

When confidence is Low, recommend a spike or research phase before committing to timeline.